O’Leary DP, Lynch N, Clancy C, Winter DC, Myers E. International, Expert-Based, Consensus Statement Regarding the Management of Acute Diverticulitis. JAMA Surg. 2015;150(9):899-904. doi:10.1001/jamasurg.2015.1675
This Delphi study provides consensus related to many aspects of acute diverticulitis and identifies other areas in need of research.
To generate an international, expert-based, consensus statement to address controversies in the management of acute diverticulitis.
Design, Setting, and Participants
This study was conducted using the Delphi technique from April 3 through October 21, 2014. A survey website was used and a panel of acute diverticulitis experts was formed via the snowball method. The top 5 acute diverticulitis experts in 5 international geographic regions were identified based on their number of publications related to acute diverticulitis.
The Delphi study used 3 rounds of questions, after which the consensus statement was collated.
Main Outcomes and Measures
A consensus statement related to the management of acute diverticulitis.
Twenty items were selected for inclusion in the consensus statement following 3 rounds of questioning. A clear definition of uncomplicated and complicated diverticulitis is provided. In uncomplicated diverticulitis, consensus was reached regarding appropriate laboratory and radiological evaluation of patients as well as nonsurgical, surgical, and follow-up strategies. A number of important topics, including antibiotic treatment, failed to reach consensus. In addition, consensus was reached regarding many nonsurgical and surgical treatment strategies in complicated diverticulitis.
Conclusions and Relevance
Controversy continues internationally regarding the management of acute diverticulitis. This study demonstrates that there is more nonconsensus among experts than consensus regarding most issues, even in the same region. It also provides insight into the status quo regarding the treatment of acute diverticulitis and provides important direction for future research.
The management of acute diverticulitis has evolved significantly during the past decade.1- 3 This evolution has been driven by a greater understanding of the disease process and has ultimately led to improved patient outcomes. However, there remain many contradictory views as to how to manage both uncomplicated and complicated episodes of acute diverticulitis.4,5
The diversity in practice paradigms appears to be caused by a lack of high-level evidence regarding best practice. This observation is reflected in international guidelines, which are mainly based on low levels of antiquated evidence. However, advances have been made in many aspects of acute diverticulitis management, including antibiotic therapy, surgical techniques, and the setting in which acute diverticulitis can be treated.6 However, embracing many of these advances into standard practice for acute diverticulitis has been slow.
Quiz Ref IDGiven this paucity of agreement with regard to the management of acute diverticulitis, we sought to achieve international consensus by conducting a Delphi study in accordance with published guidelines.7,8
A Delphi study was conducted to seek expert international opinion on the management of acute uncomplicated and complicated diverticulitis. This process took place from April 3 through October 21, 2014.
An expert panel of colorectal surgeons was formed using the snowball method. A literature search for the terms acute diverticulitis and diverticular disease was conducted using PubMed, Google Scholar, and the Cochrane databases. The top 5 published experts in 5 geographic regions, including Europe, Ireland and the United Kingdom, North America, Australia and New Zealand, and Asia were then identified. Where no reply was received from an identified expert, the next published expert in that geographic region was invited. An invitation email was sent to each expert and they were then asked to supply email addresses for 5 additional colorectal surgeons in their geographic region who they would consider an expert. A separate email was then sent to the additional experts.
The contents of the study questionnaire were generated using statements that reflected recent published guidelines. The format of the statements was edited and agreed on by senior study authors (D.P.O’L., D.C.W., and E.M.). The questionnaire was divided into 2 parts. The first part contained statements related to the management of uncomplicated diverticulitis. The second part contained statements related to the management of complicated diverticulitis. The importance of each statement was assessed using a 9-point Likert scale, with 1 indicating definitely not important; 5, undecided; and 9, definitely important.
Respondents completed the questionnaire on the SurveyMonkey website for each round of the 3-part Delphi process. In round 1, experts were invited to indicate the importance of each statement using the Likert scale and were encouraged to suggest new statements or modify existing statements. Email reminders were sent to nonresponders. Responses were analyzed and agreement was achieved when 80% or more of the participants rated criteria within 3 points on the Likert scale and the median score was 7 or greater. No consensus was considered to be achieved for statements that failed to fulfill this criterion.
Six headings comprising 42 statements made up the uncomplicated diverticulitis questionnaire. Six headings comprising 21 statements made up the complicated diverticulitis questionnaire. This method generated a total of 12 headings and 63 statements in round 1. The first round of the questionnaire was conducted from April 3 through May 29, 2014.
Only surgeons who responded to round 1 were invited to participate in round 2. Statements that failed to reach consensus with a median score of greater than 3 and less than 7 and an agreement percentage of less than 80% were re-presented in round 2. Additional suggested statements from study experts were included in round 2. Email reminders were sent to nonresponders. Round 2 was conducted from May 30 through August 11, 2014.
Only surgeons who responded to round 2 were invited to participate in round 3. Statements in round 3 were made up of edited nonconsensus statements following feedback from respondents. No further rounds were planned owing to the potential for respondent fatigue. Round 3 was conducted from August 12 through October 21, 2014.
In total, 64 surgeons were invited to participate. Fifty-four surgeons responded to round 1 (84% response rate), 30 surgeons responded to round 2 (55% response rate), and 27 surgeons responded to round 3 (90% response rate). The geographic locations of participating surgeons are presented in Table 1.
Fifteen statements had 2 levels of consensus in round 1, with a median score of 7 or greater and an agreement percentage of 80% or more (15 of 63 [23.8%]) (Table 2 and Table 3). Fifteen statements had a median score of 3 or less and an agreement percentage of less than 80% and were discarded.
Thirty-three statements had a median score of greater than 3 and less than 7 and an agreement percentage of less than 80% and were put forward for inclusion in the round 2 questionnaire. Before inclusion in round 2, all statements were assessed again, taking into account the suggestions of the experts; when necessary, new statements were generated and included.
The round 2 questionnaire consisted of 34 statements. Only 1 statement reached 1 level of consensus (median score, ≥7) but was not included in the final consensus statement because it did not reach the second level of consensus (agreement percentage, ≥80%).
The round 3 questionnaire was generated again using feedback from the experts to reach consensus regarding these topics. Ten statements were included. Five statements reached the second level of consensus (5 of 10 [50%]). One statement reached 1 level of consensus and was discarded with the remaining statements. At the end of the process, 20 items were selected for inclusion in the consensus statement (Table 4 and Table 5).
Controversies surround almost every aspect of the management of uncomplicated and complicated diverticulitis. This Delphi study demonstrates this diversity of opinion clearly by the fact that more than two-thirds of the Delphi statements ultimately failed to reach consensus. However, we do provide 20 statements whereby diverticulitis management can be practiced in line with expert opinion.
Quiz Ref IDThe setting in which acute diverticulitis is managed was addressed by this Delphi study. This topic has recently been the subject of a randomized clinical trial.9 Results from the Diver trial9 demonstrate the safety of managing acute uncomplicated diverticulitis in the outpatient setting provided a protocol is followed, including computed tomography during the assessment on initial presentation. The Delphi study results demonstrate that there is international acceptance of this approach to the management of acute uncomplicated diverticulitis.
Quiz Ref IDThis study offers a clear definition of acute uncomplicated and complicated diverticulitis. Separating acute uncomplicated diverticulitis from complicated diverticulitis relies on radiological classification and cannot be defined through clinical acumen alone. Computed tomography of the abdomen and pelvis is the most suitable modality with which to gauge the severity of acute diverticulitis; this approach was supported by our study experts. However, computed tomography is not suitable in all patients, especially in pregnant patients for whom ultrasonography should be considered. Use of intravenous contrast dye was considered important by the study experts, with no consensus reached regarding oral and rectal contrast. Little credence was given to the use of erect radiography of the chest and plain radiography of the abdomen, with experts being inclined to exclude these from the evaluation of patients with suspected acute diverticulitis if computed tomography was being conducted. In terms of laboratory values, the white blood cell count, neutrophil count, and C-reactive protein level were the only laboratory parameters the experts thought were useful.
The etiology of acute diverticulitis has traditionally been viewed as an infective process. However, the results from the AVOD Study Group6 demonstrate that there is no benefit to treating uncomplicated diverticulitis with antibiotics. This group hypothesized that uncomplicated diverticulitis represents exacerbation of a form of inflammatory bowel disease and does not necessitate antibiotic treatment. Further results from the DIABOLO trial10 are eagerly awaited. Quiz Ref IDThis Delphi study was unable to reach consensus regarding the use of a nonantimicrobial strategy in patients with acute uncomplicated diverticulitis. When we assessed the difference between geographic approaches to acute diverticulitis management, there appeared to be differences in the approach to antibiotic use. There was no consensus regarding antibiotic use in Europe, the United Kingdom, and Ireland; however, the US respondents reached a consensus in favor of the use of antibiotics for uncomplicated diverticulitis. No other statement showed any geographic difference.
Recent studies have also proposed the usefulness of adjunctive measures, such as probiotics, to prevent recurrent episodes, but these adjuncts also failed to reach expert consensus. In addition, the duration and mode of administration of antibiotics failed to reach consensus.
Surgical management of uncomplicated diverticulitis was deemed a multifactorial decision. The influence of the number of episodes of acute uncomplicated diverticulitis did not reach consensus, nor did patient age or body mass index; however, through expert feedback, it was determined that the decision whether to proceed with surgery is multifactorial and not influenced directly by any of these criteria. When surgery is considered, a laparoscopic approach should be used when appropriate.
The follow-up of patients with acute uncomplicated diverticulitis has traditionally involved a colonoscopy after the acute episode had been resolved to confirm the diagnosis. However, there is consensus that colonoscopy needs only to be performed in selected patients and may not be indicated for every patient. This consensus is again supported by recent high-level evidence.11,12
No consensus was reached among the experts regarding the use of a high-fiber diet or dietician in the follow-up of patients with acute uncomplicated diverticulitis. This lack of consensus was owing to the paucity of evidence favoring the use of fiber to prevent diverticulitis and most likely emanates from our increasing understanding regarding the underlying cause of diverticular disease and acute diverticulitis.
Quiz Ref IDThe surgical approach to complicated diverticulitis remains a contentious area. Designing statements in light of the current literature proved difficult, mainly because the choice of surgical approach depends on the degree of sepsis present and must be tailored to each patient. The experts agreed with this viewpoint that the choice of surgical approach is mainly indicated by the degree of sepsis. The only surgical approach that achieved consensus was the use of the Hartmann procedure in the setting of feculent peritonitis.
Although this study was directed by statements from publications and opinions from experts, there are limitations to the Delphi technique. When consensus is reached, it is based on opinion only. Also, similar to any published guidelines, they are out of date almost as soon as they are published; however, it is hoped that this study can be a useful tool for directing future research. Specific to our Delphi study, the response rate was excellent for the first round of the study but decreased in round 2. Feedback from the experts who did not participate in later rounds responded that there were too many statements and participation was time consuming. Finally, we aimed to encompass all practice parameters pertaining to acute diverticulitis; however, owing to respondent fatigue, we were unable to include any aspect of diverticular fistula management.
Controversy continues internationally regarding the management of acute diverticulitis. This study demonstrates that there is more nonconsensus among experts than consensus, even in the same region, on most issues. It also provides an excellent snapshot of the status quo in the treatment of acute diverticulitis as well as important directions for future research.
Accepted for Publication: March 22, 2015.
Corresponding Author: D. Peter O’Leary, PhD, Department of Colorectal Surgery, Galway University Hospital, Galway, Ireland (firstname.lastname@example.org).
Published Online: July 15, 2015. doi:10.1001/jamasurg.2015.1675.
Author Contributions: Dr Myers had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: O’Leary, Lynch, Winter, Myers.
Acquisition, analysis, or interpretation of data: O’Leary, Lynch, Clancy, Winter.
Drafting of the manuscript: O’Leary, Lynch, Clancy, Myers.
Critical revision of the manuscript for important intellectual content: O’Leary, Lynch, Winter.
Statistical analysis: O’Leary.
Administrative, technical, or material support: O’Leary, Lynch, Clancy.
Study supervision: Winter, Myers.
Conflict of Interest Disclosures: None reported.